The present invention relates to a data processing method and apparatus, and a recording medium for an error-correcting product code favorable for use in the recording and transmission of digital data.
More particularly, the present invention relates to a data processing system using an error-correcting product code which comprises an error-correcting outer parity and an error-correcting inner parity which are effective in the case where information data is recorded on a plurality kinds of recording media particularly having a largely different recording density. Here, particularly, in a method for forming the outer parity, a PO series creation by n sets of data items aggregated by n rows is used. Consequently, even when the error-correcting product code block is recorded on a recording medium in an order of data transmission without carrying out data interleave process; the capability of coping with the defect is largely improved.
In a system in which digital data is recorded on an optical disk by bytes (one byte is equal to eight bits) or digital data is transmitted to a transmission channel, a Reed-Solomon error-correcting product code block is constructed to process data. That is, (M.times.N) bytes of data is arranged in a matrix containing an M rows.times.N columns. Then, PO bytes error-correcting word is added to each column of M bytes information portion. Then, PI bytes of error-correcting words are added to each row of N bytes information portion. Then, (M+Po) rows.times.(N+Pi) columns Reed-Solomon error-correcting product code block is constructed. Then, this Reed-Solomon error-correcting product code block is either recorded on a recording medium or transmitted to a transmission channel. The error-correcting processing portion on the information reproduction side of the recording medium and the receiving side of the transmission channel are capable of correcting random errors and burst errors on the information portion by using the error-correcting words PO and PI.
Such Reed-Solomon error-correcting product code block has a higher data processing efficiency with a decrease in a ratio of a redundant portion (Pi.times.M+Po.times.N+Po.times.Pi) of the error-correcting word with respect to the whole word referred to as redundancy ratio, namely (M+Po).times.(N+Pi). On the other hand, the error-correcting capability is also heightened with respect to the random error and the burst error with an increase in the Pi and Po.
Here, it is known that the Reed-Solomon error-correcting code block having small M and N, namely small Pi and Po has a lower correcting capability because of relatively higher probability of error in error correction in the case where the Reed-Solomon error-correcting product code blocks having the same redundancy ratio are compared with each other.
On the contrary, it is known that since Pi and Po can be increased at the same redundancy ratio with an increase in M and N, a high error-correcting capability can be obtained. However, such capability cannot be realized unless the constraint conditions described below are satisfied.
A first constraint condition is that M+Po and M+Pi must be equal to or less than 255 bytes as a code length for constructing the Reed-Solomon error-correcting product code block (in the case where the length of the code is eight bits). Incidentally, Pi described above refers to the PI series error-correcting code length while Po refers to the PO series error-correcting code length.
A second constraint condition is a cost constraint resulting from the scale of the hardware.
By the way, when considered on the basis of the above conditions, optical disk standards such as a DVD-ROM, a DVD-RAM, a DVD-R or the like which are the information recording media in recent years are made public as a standard in which the improved Reed-Solomon error-correcting product code block is adopted. Out of these standards, the DVD-ROM and the DVD-RAM are established as DIS16448 (DVD-ROM having a diameter of 80 mm) and DIS16449 (DVD-ROM having a diameter of 120 mm) and DIS16825 (DVD-RAM).
In this DVD standard, the above idea is adopted with respect to the error-correcting word processing method so that the error-correcting capability is remarkably improved with error-correcting word having a small redundancy ratio as compared with the method used in the conventional optical disks.
The concept on the error-correcting method of the DVD is basically described above, the fundamental problem is to what level the target of the random error-correcting capability and the burst error-correcting capability is to be set. In order to set such level, the recording method of the recording medium and the generation of defects resulting from the handling thereof must be considered.
The recording/reproducing method is determined from the recording/reproducing beam spot size resulting from the recording wavelength and the optical system characteristic in the optical disk system. Here, the recording density constitutes a large factor in the determination of the error-correcting method. In particular, in the determination of the burst error correction capability, the defect length such as scratches or the like generated in the handling of the discs can be determined from experience. With respect to the error-correcting capability, the multiplication of line recording density by the physical defect length constitutes a burst error length of information data with the result that the error correcting capability is required to be raised in the improvement of the recording density.
The recording density can be described as follows with particular reference to the reproduction system.
When, a light source wavelength is denoted by .lambda., and a numeric aperture of an object lens is denoted by NA, the recording density stands proportional to (NA/.lambda.).sup.2. The wavelength adopted in the DVD is 650 nm while NA is 0.6.
In the error-correcting method, a row side inner parity of RS (182, 172, 11) and a column side outer parity of RS (208, 192, 17) are adopted by means of PI (inner parity)=10 bytes and PO (outer parity)=16 respectively with respect to the (M.times.N)=(192.times.172) bytes information data block in terms of the Reed-Solomon error-correcting product code (RS is referred to as Reed-Solomon). The block used in this error-correcting method is referred to as the error-correcting product code block.
Here, with respect to the error-correcting product code block, the error is corrected in the PI series at first, and an error mark is attached to a row whose error cannot be corrected. Thereafter, at the time of the error correction on the PO series, the error mark is treated as an error position. When the so-called “erasure correction” method for calculating and extracting only error patterns is used, a maximum of 16 rows of burst errors can be corrected. In the DVD, since the recording density is data bit length=0.267.mu.m, 0.000267.times.8.times.182.times.-16=6.2 mm is given. It is possible to say that about 6 mm burst error-correcting capability is given.
However, as a next generation DVD an examination is started on an optical disk having a large capacity resulting from further increase in the density. For the increase in capacity exceeding the DVD, the recording density must be increased. Recently, in order to meet such request, a blue laser diode having a wavelength of 450 nm is made public. When such laser diode is used, it is expected that the recording density can be improved by about 2.6 times in the optical system similar to the DVD or the like. With the improvement in the optical system, four to five times higher density can be realized so that a fine image such as a high definition image such as a Hi-Vision or the like can be recorded for two or more hours on one disc.
In such increase in the density (for example, the line density is about twice as compared with the conventional one), only about 3 mm error-correcting capability can be provided with respect to the burst error when the conventional error-correcting method is introduced.
Furthermore, as described above, the error-correcting word length is 255 bytes at most as long as 1 word=8 bit system processing system is used. Since the PO series is 208 bytes in the DVD standard, the burst error correcting capability is close to the limit in the above error-correcting method so that only little improvement can be expected.
In order to expand the error-correcting word length, the word length may only be lengthened. With respect to the word length, a multiple of eight can be easily used. As a consequence, 1 word=16 bits can be considered. The scale of the error-correcting circuit as hardware is extremely large as compared with the conventional one so that there arises many problems.
In such a case, there is generally available a technique in which the burst error-correcting capability is improved while maintaining the error-correcting code length by adopting a data interleave to scatter the burst error.
However, the data interleave is not adopted even in the DVD standard. The reason goes as follows: in the case where an error is created which exceeds the error-correcting capability in the reproduction processing in the case of an image signal in which information data is compressed, the error data is scattered with the result that a disadvantage of the reproduced image is generated at many positions. In the reproduction processing of the image signal, it is thought that processing of concentrating and reproducing disadvantageous images as much as possible is favorable as a processing of generated disadvantage. This is because the processing can be completed with the reproduction of an instant disadvantageous image.
Besides, a structure close to the current DVD system is favorable for the upper compatibility with respect to the next generation system.
Points Noted by the Inventors of the Present Invention
Generally, in the error-correcting processing method such as a package medium or the like, the Reed-Solomon product code block method is introduced in many cases. This is because high performance and high efficiency can be expected with this method in the case where defect error data such as defects generated in package medium or the like is detected and corrected.
With respect to the unit of processed data, 1 word=1 byte (8 bits) is favorable. When the application development of the system is considered, it is required to suppress a processing circuit to an appropriate hardware scale. Besides, this fact is required for facilitating a connection to the recording medium and the transmission channel because a front and rear processing circuit is provided in the recording on the recording medium and data transmission to the transmission channel in addition to the error-correcting processing.
Under these circumstances, use of the Reed-Solomon error-correcting product code block used in the current DVD is optimal as an error-correcting method which can corresponds to a large improvement in the recording density of the recording medium under the above surrounding situation.
row side inner parity RS (182, 172, 11)
column side outer parity RS (208, 192 17)
Here, the problem is that it is required to settle the improvement of the burst error correcting capability.
In order to heighten the burst error correcting capability, the error may only be scattered in the error detecting and correcting capability in each of the correction code. However, an image and an acoustic signal as information data are subject to compression coding. In a system for recording and reproducing the compression signal, a data structure or error-correcting processing system is desirable which is capable of suppressing information breakdown to a minimal level in the final reproduction of the image and acoustic signal.
In particular, as countermeasures for dealing with the burst error, the number of errors in one error correction block is decreased by scattering the error signal, so that the error-correcting capability can be improved. However, in the case where errors are present in the number exceeding the error-correcting capability, the dispersion of the error signal will result in the expansion of the damage done on the whole data. Consequently, it is difficult to adopt the method using the error data dispersion, namely, the data interleave, which constitutes the basic concept of heightening the burst error-correcting capability.